Ataxia RNAseq - Leafcutter junction category distributions

UCL - Version 1.3

1 Background

Aim: update the studies carried out by Regina H. Reynolds about introns’ annotation distributions. Original source material: Leafcutter: intron annotation and visualisation

This report is meant to be used as a template. As such, the different sections will include instructions on how to modify the graphic outputs and not descriptions about the results themselves. To run this script, it is necessary to source the file at /home/grocamora/Core_Projects/Ataxia_Splicing_Analysis/R/Leafcutter_introns.R with needed helper functions.

Changelog

v1.3

• Changed the reference transcriptome to Gencode v38.

v1.2

• Changed the reference transcriptome to Gencode v39.

2 Annotate introns

To annotate the junctions, we need to set the following variables:

• use_strand: whether to use the strand information for the Leafcutter clusters’ ID. Note that the script was not tested with use_strand set to TRUE, so it might contain errors.

• gtf_path: path to the reference transcriptome to annotate the junctions. In this report, GENCODE v38 is employed (also tested with ENSEMBL v105).

# For the Ataxia dataset, the cluster IDs don't have a valid strand. Set to FALSE.
use_strand <- F

# Path to the gtf file to annotate the introns
#gtf_path <- "/home/grocamora/RytenLab-Research/Additional_files/Homo_sapiens.GRCh38.105.gtf"
gtf_path <- "/home/grocamora/RytenLab-Research/Additional_files/GENCODE/gencode.v38.annotation.gtf"

2.1 Loading leafcutter results

Next, the leafcutter results are loaded from disk. Three different studies will be presented in this report: overall (Level 1), subtypes (Level 2) and diagnoses (Level 3). Types 2 and 3 will be further split by tissues.

path_to_results <- file.path(main_path, "data/leafcutter_results/")
leafcutter_list <-  setNames(list(
  readRDS(file.path(path_to_results, "case_v_control_leafcutter_ds.Rds")),
  readRDS(file.path(path_to_results, "subtypes_v_control_leafcutter_ds.Rds")),
  readRDS(file.path(path_to_results, "diagnoses_v_control_leafcutter_ds.Rds"))),
  c("overall", "subtypes", "diagnoses"))

2.2 Annotation process

The junctions are annotated using the function junction_annot from David Zhang’s package dasper version 1.7.0.

# Path to where the annotated intron list will be stored in disk.
annotated_path <- file.path(main_path, "variables/annotated_gencode38_test.rds")

# If file already exists, do not execute the annotation process.
if(!file.exists(annotated_path)){
  # Load the reference transcriptome
  TxDb_ref <- dasper:::ref_load(gtf_path)
  
  # Some reference transcriptomes (i.e. Gencode) use the characters "chr" before
  # the seqnames. To bring consistency, the "seqlevelsStyle" is set to "NCBI".
  #GenomeInfoDb::seqlevelsStyle(TxDb_ref) <- c("NCBI")
  
  annotated <- leafcutter_list %>% purrr::map(function(x){
    # Extract the successful clusters
    success_cluster <- x$cluster_significance %>%
      dplyr::filter(status == "Success") %>% 
      dplyr::pull(cluster) %>% 
      unique()
    
    # Transform the data from leafcutter to a valid GRanges object to run the annotation
    annotated_introns <- x$intron_usage %>%
      tidyr::separate(col = "intron", into = c("chr", "intron_start", "intron_end", "cluster_id"), sep = ":", remove = F) %>%
      dplyr::mutate(cluster_id = str_c(chr, ":", cluster_id)) %>%
      dplyr::filter(cluster_id %in% success_cluster) %>%
      dplyr::select(-chr, -intron_start, -intron_end, cluster_id) %>%
      convert_leafcutter(use_strand = use_strand, seqlevelsStyle = GenomeInfoDb::seqlevelsStyle(TxDb_ref)) %>%
      dasper::junction_annot(ref = TxDb_ref)
    
    # Once the introns are annotated, we remove all the annotated introns that
    # are associated to more than one gene.
    annotated_introns <- removeAmbiguousGenesLeafcutter(annotated_introns)
    
    return(annotated_introns)
  })
  
  # Save the annotated list in disk.
  annotated %>% saveRDS(annotated_path)
}else{
  annotated <- readRDS(annotated_path)
}

It is important to use a consistent transcriptome seqlevelsStyle in order to properly match the extracted junctions. By default, NCBI style is recommended (no chr before the seqnames).

2.3 Proportion of annotated introns

We use the function printTypeTable to print a table with the proportion of significant junctions by category. Results by tissue are shown when expanded below the main table. The significant column denotes the number of introns by annotation type that were found significantly differentially spliced. Original Source.

#' Prints the proportion of significant junctions by type
#'
#' Given the list of annotated junctions and the list of leafcutter results, this
#' function extract a given index of the lists (corresponding to different
#' studies) and prints a table with the proportion of significantly
#' differentially spliced junction by type.
#'
#' @param annotated list of annotated junctions. Each element corresponds to a specific study.
#' @param leafcutter_list list of leafcutter results. Each element corresponds to a specific study.
#' @param level index or name of the specific study in both the "annotated" and "leafcutter_list" arguments.
#' @param use_strand boolean to specify whether to use the strand in the clusters' ID.
#' @param deltapsi_filter boolean to specify whether to use a deltaPSI filter of |dPSI| >= 0.1
#' @param tissue character vector, if provided, the comparisons employed must contain the keyword.
#'
#' @export
printTypeTable <- function(annotated, leafcutter_list, level, use_strand = F, deltapsi_filter = F, tissue = NULL){
  # Extract the significant comparisons and clusters
  significant_clusters <- leafcutter_list[[level]]$significant_clusters_0.05_filter %>%
    dplyr::mutate(cluster_id = str_c(chr, ":", cluster)) %>% 
    removeClusterStrand(use_strand, input_name = "cluster_id") %>%
    {if(deltapsi_filter) dplyr::filter(., abs(deltapsi) >= 0.1) else .} %>%
    dplyr::distinct(comparison, cluster_id)
  
  # Convert the annotated GRanges object to a tibble.
  annotated_df <- annotated[[level]] %>%
    tibble::as_tibble()
  
  # If a tissue is provided, filter the comparisons to contain the input tissue.
  if(!is.null(tissue)){
    significant_clusters <- significant_clusters %>%
      dplyr::filter(grepl(tissue, comparison))
    annotated_df <- annotated_df %>%
      dplyr::filter(grepl(tissue, comparison))
  }
  
  # Extract the number of junctions by type that are significant.
  significant_counts <- annotated_df %>%
    dplyr::inner_join(significant_clusters) %>%
    dplyr::group_by(type) %>% # (Optional) Add comparison/tissue here.
    dplyr::summarise(significant = n())
  
  # Prints the proportion of significant junctions by type.
  annotated_df %>%
    dplyr::group_by(type) %>% # (Optional) Add comparison/tissue here.
    dplyr::summarise(all = n()) %>% 
    dplyr::inner_join(significant_counts) %>%
    dplyr::mutate(proportion = (significant/all) %>% signif(2)) %>% 
    dplyr::arrange(-all) %>% 
    `colnames<-`(c("Junction Type", "Count Junctions", "Significant Junctions", "Proportion")) %>%
    kableExtra::kbl(booktabs = T, linesep = "") %>%
    kableExtra::kable_classic(full_width = T, "hover", "striped", html_font = "Cambria", font_size = 14) %>%
    kableExtra::row_spec(0, bold = T, font_size = 16)
}

Level 1 (Type)

Lenient

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	216286	5826	0.027
novel_acceptor	45699	998	0.022
novel_donor	37020	966	0.026
novel_exon_skip	28717	827	0.029
unannotated	12056	171	0.014
novel_combo	3679	141	0.038

Information by tissue

Cerebellum:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	104917	2588	0.025
novel_acceptor	22637	439	0.019
novel_donor	18174	378	0.021
novel_exon_skip	14332	384	0.027
unannotated	6278	103	0.016
novel_combo	1767	52	0.029

Frontal Cortex:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	111369	3238	0.029
novel_acceptor	23062	559	0.024
novel_donor	18846	588	0.031
novel_exon_skip	14385	443	0.031
unannotated	5778	68	0.012
novel_combo	1912	89	0.047

Stringent

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	216286	1078	0.0050
novel_acceptor	45699	174	0.0038
novel_donor	37020	141	0.0038
novel_exon_skip	28717	72	0.0025
unannotated	12056	55	0.0046
novel_combo	3679	22	0.0060

Information by tissue

Cerebellum:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	104917	583	0.0056
novel_acceptor	22637	90	0.0040
novel_donor	18174	75	0.0041
novel_exon_skip	14332	43	0.0030
unannotated	6278	39	0.0062
novel_combo	1767	11	0.0062

Frontal Cortex:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	111369	495	0.0044
novel_acceptor	23062	84	0.0036
novel_donor	18846	66	0.0035
novel_exon_skip	14385	29	0.0020
unannotated	5778	16	0.0028
novel_combo	1912	11	0.0058

Level 2 (AtaxiaSubtype)

Lenient

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	590780	11034	0.019
novel_acceptor	114597	1998	0.017
novel_donor	91902	1828	0.020
novel_exon_skip	75288	1545	0.021
unannotated	25133	358	0.014
novel_combo	9366	290	0.031

Information by tissue

Cerebellum:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	276558	4180	0.015
novel_acceptor	54709	824	0.015
novel_donor	42953	707	0.016
novel_exon_skip	37093	655	0.018
unannotated	12398	169	0.014
novel_combo	4275	113	0.026

Frontal Cortex:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	314222	6854	0.022
novel_acceptor	59888	1174	0.020
novel_donor	48949	1121	0.023
novel_exon_skip	38195	890	0.023
unannotated	12735	189	0.015
novel_combo	5091	177	0.035

Stringent

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	590780	3088	0.0052
novel_acceptor	114597	495	0.0043
novel_donor	91902	480	0.0052
novel_exon_skip	75288	263	0.0035
unannotated	25133	180	0.0072
novel_combo	9366	88	0.0094

Information by tissue

Cerebellum:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	276558	1249	0.0045
novel_acceptor	54709	221	0.0040
novel_donor	42953	196	0.0046
novel_exon_skip	37093	125	0.0034
unannotated	12398	90	0.0073
novel_combo	4275	42	0.0098

Frontal Cortex:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	314222	1839	0.0059
novel_acceptor	59888	274	0.0046
novel_donor	48949	284	0.0058
novel_exon_skip	38195	138	0.0036
unannotated	12735	90	0.0071
novel_combo	5091	46	0.0090

Level 3 (Diagnosis)

Lenient

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	600885	50236	0.084
novel_acceptor	121969	8637	0.071
novel_donor	94296	7159	0.076
novel_exon_skip	86016	8078	0.094
unannotated	22513	1212	0.054
novel_combo	9910	1209	0.120

Information by tissue

Cerebellum:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	237320	26248	0.110
novel_acceptor	48677	4463	0.092
novel_exon_skip	36791	4861	0.130
novel_donor	35990	3490	0.097
unannotated	9644	600	0.062
novel_combo	3844	502	0.130

Frontal Cortex:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	363565	23988	0.066
novel_acceptor	73292	4174	0.057
novel_donor	58306	3669	0.063
novel_exon_skip	49225	3217	0.065
unannotated	12869	612	0.048
novel_combo	6066	707	0.120

Stringent

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	600885	23016	0.038
novel_acceptor	121969	3927	0.032
novel_donor	94296	3266	0.035
novel_exon_skip	86016	3077	0.036
unannotated	22513	702	0.031
novel_combo	9910	626	0.063

Information by tissue

Cerebellum:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	237320	12509	0.053
novel_acceptor	48677	2130	0.044
novel_exon_skip	36791	2052	0.056
novel_donor	35990	1702	0.047
unannotated	9644	399	0.041
novel_combo	3844	279	0.073

Frontal Cortex:

Junction Type	Count Junctions	Significant Junctions	Proportion
annotated	363565	10507	0.029
novel_acceptor	73292	1797	0.025
novel_donor	58306	1564	0.027
novel_exon_skip	49225	1025	0.021
unannotated	12869	303	0.024
novel_combo	6066	347	0.057

3 Visualization of overlapping events

Different visualizations will be presented replicating Regina’s results, but adapting the code to remove depecrated functions and adding the option to use non-stranded clusters.

3.1 Proportion of annotation types within each comparison

Visualization of the proportion of junction categories by comparison and tissue. We use the function plotProportionOfAnnotation with the parameters explained next. Relevant to notice the addition of the tissue parameter that can split the comparisons by keywords in their name (e.g. if we specify Cerebellum in the tissue parameter, only comparison that include the keyword Cerebellum will be considered). Original Source

#' Plots the proportion of junction categories
#'
#' Given the list of annotated junctions and the list of leafcutter results,
#' this function extract a given index of the lists (corresponding to different
#' studies) and plots the proportion of each category.
#'
#' Additional parameters include the option to use the cluster strand or to
#' divide by tissue.
#'
#' @param annotated list of annotated junctions. Each element corresponds to a
#'   specific study.
#' @param leafcutter_list list of leafcutter results. Each element corresponds
#'   to a specific study.
#' @param level index or name of the specific study in both the "annotated" and "leafcutter_list" arguments.
#' @param use_strand boolean to specify whether to use the strand in the
#'   clusters' ID.
#' @param tissue character vector with the specific tissue to study. The tissue
#'   name must be contained in the comparison name.
#'
#' @return
plotProportionOfAnnotation <- function(annotated, leafcutter_list, level, use_strand = F, tissue = NULL){
  # Get the successful comparisons and clusters
  success_cluster <- leafcutter_list[[level]]$cluster_significance %>%
    # If tissue is provided, it filter the comparisons by tissue
    {if(!is.null(tissue)) dplyr::filter(., grepl(tissue, comparison)) else .} %>%
    dplyr::filter(status == "Success") %>%
    dplyr::distinct(comparison, cluster) %>%
    removeClusterStrand(use_strand)
  
  # Extract the introns with successful comparison and cluster
  success <- annotated[[level]] %>%
    tibble::as_tibble() %>%
    dplyr::inner_join(success_cluster)
  
  # Get introns with significant comparison and cluster
  significant <- extractSignificantIntrons(annotated[[level]], leafcutter_list[[level]], use_strand, tissue)
  
  # Plots will be stored in a list 
  intron_list <- list(success, significant)
  plot_list <- vector(mode = "list", 2)
  
  for(j in 1:length(intron_list)){
    # Calculate the proportion of each category by comparison
    count <- intron_list[[j]] %>% 
      dplyr::filter(!type == "ambig_gene") %>% 
      dplyr::group_by(comparison, type) %>% 
      dplyr::summarise(n = n()) %>% 
      dplyr::mutate(prop = n/sum(n)) %>% 
      dplyr::ungroup()
    
    plot_list[[j]] <- count %>% 
      dplyr::mutate(type = type %>% factor(levels = intron_type_levels %>% rev()),
                    #comparison = comparison_labels[comparison],
                    comparison = as_factor(comparison)) %>% 
      ggplot(aes(x = comparison, y = prop, fill = type), colour = "black") +
      geom_col(color = "black", width= 0.6) +
      # ggrepel::geom_text_repel(aes(label = paste0(round(prop, 2)*100, "%"), x = comparison, group = type), 
      #                          position = ggpp::position_stacknudge(x = 0.4, direction = "split"),
      #                          size = 3.5) +
      # ggrepel::geom_text_repel(aes(label = round(prop, 2), x = comparison, group = type), direction = "x", force= 10) +
      labs(x = "", y = "Proportion") +
      scale_y_continuous(expand = expansion(mult = c(0, 0.05))) + 
      scale_x_discrete(expand = expansion(add = 0.5)) + 
      scale_fill_manual(name = "Acceptor/donor annotation", 
                        values = rev(c("#3C5488", "#E64B35", "#00A087", "#4DBBD5", "#7E6148", "grey"))) +
      custom_gg_theme + 
      theme(axis.text.x = ggplot2::element_text(color = "black", size = 8, angle = 90, hjust = 0.5)) +
      guides(fill = guide_legend(reverse = T))
  }
  
  # Arrange both plots in two columns
  ggpubr::ggarrange(plotlist = plot_list,
                    labels = c("a", "b"),
                    common.legend = TRUE, legend = "top") %>%
    ## If tissue is provided, add a title (remove next line to ignore title)
    {if(!is.null(tissue)) ggpubr::annotate_figure(., top = ggpubr::text_grob(paste0("Tissue: ", tissue), size = 16)) else .}
}

Level 1 (Type)

Figure 3.1: Proportion of (a) all successfully tested introns and (b) the subset of differentially spliced introns coloured by annotation type.

Level 2 (AtaxiaSubtype)

Cerebellum

Figure 3.2: Proportion of (a) all successfully tested introns and (b) the subset of differentially spliced introns coloured by annotation type. Results only for Cerebellum tissue.

Frontal Cortex

Figure 3.3: Proportion of (a) all successfully tested introns and (b) the subset of differentially spliced introns coloured by annotation type. Results only for Frontal Cortex tissue.

Level 3 (Diagnosis)

Cerebellum

Figure 3.4: Proportion of (a) all successfully tested introns and (b) the subset of differentially spliced introns coloured by annotation type. Results only for Cerebellum tissue.

Frontal Cortex

Figure 3.5: Proportion of (a) all successfully tested introns and (b) the subset of differentially spliced introns coloured by annotation type. Results only for Frontal Cortex tissue.

3.2 Number of overlapping introns

In the following plots, we measure the number of differentially spliced introns across overlaps. We use the function plotNumberOverlappingIntrons, with the same arguments as before. Original Source

#' Plot the number of overlapping significant junctions
#'
#' First, we obtain the annotated introns from significant clusters and
#' comparisons. Then, we remove the junction category "ambig_gene" and count the
#' number of entries by intron ID and category.
#'
#' @param annotated list of annotated junctions. Each element corresponds to a
#'   specific study.
#' @param leafcutter_list list of leafcutter results. Each element corresponds
#'   to a specific study.
#' @param level index or name of the specific study in both the "annotated" and "leafcutter_list" arguments.
#' @param use_strand boolean to specify whether to use the strand in the
#'   clusters' ID.
#' @param tissue character vector with the specific tissue to study. The tissue
#'   name must be contained in the comparison name.
#'
#' @return
#' @export
plotNumberOverlappingIntrons <- function(annotated, leafcutter_list, level, use_strand = T, tissue = NULL){
  # Get introns with significant comparison and cluster
  significant <- extractSignificantIntrons(annotated[[level]], leafcutter_list[[level]], use_strand, tissue)
  
  # Find number of overlapped introns between the comparisons
  number_ovelaps <- significant %>% 
    dplyr::filter(!type == "ambig_gene") %>% 
    dplyr::mutate(unique_id = str_c(cluster_id, ":", start, ":", end),
                  type = type %>% factor(levels = intron_type_levels)) %>% 
    dplyr::group_by(unique_id, type) %>% 
    dplyr::mutate(n_overlaps = n()) %>%
    dplyr::ungroup() %>%
    dplyr::mutate(n_overlaps = factor(n_overlaps, levels = 1:length(unique(comparison)))) %>%
    dplyr::distinct(unique_id, type, n_overlaps)
  
  # Plot the results
  ggplot(number_ovelaps, aes(x = n_overlaps, fill = type)) +
    geom_bar(position = position_dodge2(preserve = "single", padding = 0), color = "black") +
    facet_wrap(~ type, labeller = labeller(type = intron_type_labels)) +
    labs(x = "Number of overlaps") +
    scale_y_continuous(expand = expansion(mult = c(0, 0.05))) + 
    scale_fill_manual(name = "Acceptor/donor annotation", 
                      values = c("#3C5488", "#E64B35", "#00A087", "#4DBBD5", "#7E6148", "grey"),
                      labels = intron_type_labels) +
    custom_gg_theme +
    theme(axis.text.x = ggplot2::element_text(color = "black", size = 8, angle = 0, hjust = 0.5)) +
    ## If tissue is provided, add a title (remove next line to ignore title)
    {if(!is.null(tissue)) ggtitle(paste0("Tissue: ", tissue))}
}

Level 1 (Type)

Figure 3.6: Number of differentially spliced introns across overlaps.

Level 2 (AtaxiaSubtype)

Cerebellum

Figure 3.7: Number of differentially spliced introns across overlaps.

Frontal Cortex

Figure 3.8: Number of differentially spliced introns across overlaps.

Level 3 (Diagnosis)

Cerebellum

Figure 3.9: Number of differentially spliced introns across overlaps.

Frontal Cortex

Figure 3.10: Number of differentially spliced introns across overlaps.

3.3 Distribution of annotation types across comparisons

We can also plot by comparisons the proportion of the number of overlapping introns. Original Source.

#' Plot the proportion of overlapping significant junctions by comparison
#'
#' First, we obtain the annotated introns from significant clusters and
#' comparisons. Then, we remove the junction category "ambig_gene" and count the
#' number of entries by intron ID and category. The proportion of each category
#' is calculated for each number of overlaps.
#'
#' @param annotated list of annotated junctions. Each element corresponds to a
#'   specific study.
#' @param leafcutter_list list of leafcutter results. Each element corresponds
#'   to a specific study.
#' @param level index or name of the specific study in both the "annotated" and "leafcutter_list" arguments.
#' @param use_strand boolean to specify whether to use the strand in the
#'   clusters' ID.
#' @param tissue character vector with the specific tissue to study. The tissue
#'   name must be contained in the comparison name.
#'
#' @return
#' @export
plotDistributionAnnotatedTypes <- function(annotated, leafcutter_list, level, use_strand = F, tissue = NULL){
  # Get introns with significant comparison and cluster
  significant <- extractSignificantIntrons(annotated[[level]], leafcutter_list[[level]], use_strand, tissue)
  
  # Find number of overlapped introns between the comparisons
  number_ovelaps <- significant %>% 
    dplyr::filter(!type == "ambig_gene") %>% 
    dplyr::mutate(unique_id = str_c(cluster_id, ":", start, ":", end),
                  type = type %>% factor(levels = intron_type_levels)) %>% 
    dplyr::group_by(unique_id, type) %>% 
    dplyr::mutate(n_overlaps = n() %>% as_factor()) %>%
    dplyr::ungroup() %>%
    dplyr::mutate(n_overlaps = factor(n_overlaps, levels = 1:length(unique(comparison))))
  
  # Find the proportion of each category for a given number of overlaps
  number_overlaps_proportion <- number_ovelaps %>% 
    dplyr::distinct(comparison, unique_id, n_overlaps, type) %>% 
    dplyr::group_by(comparison, n_overlaps, type) %>% 
    dplyr::summarise(n = n()) %>% 
    dplyr::group_by(comparison) %>% 
    dplyr::mutate(prop = n/sum(n))
  
  # Plot the results
  ggplot(number_overlaps_proportion,
         aes(x = n_overlaps, y = prop, fill = type)) +
    geom_col(position = position_dodge2(preserve = "single", padding = 0), color = "black") +
    facet_grid(type ~ comparison, scales = "free_y", labeller = labeller(type = intron_type_labels)) +
    labs(y = "Proportion of differentially spliced introns") +
    scale_y_continuous(expand = expansion(mult = c(0, 0.05))) + 
    scale_fill_manual(name = "Acceptor/donor annotation", 
                      values = c("#3C5488", "#E64B35", "#00A087", "#4DBBD5", "#7E6148", "grey"),
                      labels = intron_type_labels) +
    custom_gg_theme +
    theme(axis.text.x = ggplot2::element_text(color = "black", size = 8, angle = 0, hjust = 0.5)) +
    {if(!is.null(tissue)) ggtitle(paste0("Tissue: ", tissue))}
}

Level 1 (Type)

Figure 3.11: Proportion of differentially spliced introns per overlap across annotation types and comparisons. Proportions were calculated per comparison, by dividing the number of distinct events in each annotation type by the total number across a comparison.

Level 2 (AtaxiaSubtype)

Cerebellum

Figure 3.12: Proportion of differentially spliced introns per overlap across annotation types and comparisons. Proportions were calculated per comparison, by dividing the number of distinct events in each annotation type by the total number across a comparison.

Frontal Cortex

Figure 3.13: Proportion of differentially spliced introns per overlap across annotation types and comparisons. Proportions were calculated per comparison, by dividing the number of distinct events in each annotation type by the total number across a comparison.

Level 3 (Diagnosis)

Cerebellum

Figure 3.14: Proportion of differentially spliced introns per overlap across annotation types and comparisons. Proportions were calculated per comparison, by dividing the number of distinct events in each annotation type by the total number across a comparison.

Frontal Cortex

Figure 3.15: Proportion of differentially spliced introns per overlap across annotation types and comparisons. Proportions were calculated per comparison, by dividing the number of distinct events in each annotation type by the total number across a comparison.

3.4 Proportion of clusters containing only annotated introns

Lastly, we also represent the proportion of significant cluster in which all the introns are annotated introns. Original Source

#' Plot the proportion significant clusters containing only annotated introns
#'
#' Obtain the introns from significant clusters and comparisons and removes the
#' junction category "ambig_gene". The proportion of annotated introns in each
#' cluster and comparison is calculated and represented.
#'
#' @param annotated list of annotated junctions. Each element corresponds to a
#'   specific study.
#' @param leafcutter_list list of leafcutter results. Each element corresponds
#'   to a specific study.
#' @param level index or name of the specific study in both the "annotated" and "leafcutter_list" arguments.
#' @param use_strand boolean to specify whether to use the strand in the
#'   clusters' ID.
#' @param tissue character vector with the specific tissue to study. The tissue
#'   name must be contained in the comparison name.
#'
#' @return
#' @export
plotDistributionOnlyAnnotatedTypes <- function(annotated, leafcutter_list, level, use_strand = F, tissue = NULL){
  # Get introns with significant comparison and cluster
  significant <- extractSignificantIntrons(annotated[[level]], leafcutter_list[[level]], use_strand, tissue)
  
  # Find number of overlapped introns between the comparisons
  number_ovelaps <- significant %>% 
    dplyr::filter(!type == "ambig_gene") %>% 
    dplyr::mutate(unique_id = str_c(cluster_id, ":", start, ":", end),
                  type = type %>% factor(levels = intron_type_levels)) %>% 
    dplyr::group_by(unique_id, type) %>% 
    dplyr::mutate(n_overlaps = n() %>% as_factor()) %>%
    dplyr::ungroup()
  
  # Count the proportion of clusters with all their introns categorized as
  # annotated
  cluster_annot_count <- number_ovelaps %>% 
    dplyr::group_by(comparison, cluster_id) %>% 
    dplyr::summarise(introns_all_annotated = all(in_ref)) %>% 
    dplyr::group_by(comparison, introns_all_annotated) %>% 
    dplyr::summarise(n = n()) %>% 
    dplyr::mutate(prop = n/sum(n)) %>% 
    dplyr::ungroup()
  
  # Plot the results
  cluster_annot_count %>% 
    dplyr::mutate(comparison = as_factor(comparison)) %>% 
    ggplot(aes(x = comparison, y = prop, fill = introns_all_annotated)) +
    geom_col(color = "black") +
    geom_text(aes(label = paste0(round(prop, 2)*100, "%")), position = position_stack(vjust = 0.5), 
              size = 4, color = "black") +
    scale_y_continuous(expand = expansion(mult = c(0, 0.05))) + 
    scale_fill_manual(name = "All introns annotated", 
                      values = c("#E64B35", "#3C5488")) +
    labs(y = "Proportion of clusters") +
    custom_gg_theme +
    {if(!is.null(tissue)) ggtitle(paste0("Tissue: ", tissue))}
}

Level 1 (Type)

Figure 3.16: Proportion of differentially spliced clusters containing introns that are all fully annotated.

Level 2 (AtaxiaSubtype)

Figure 3.17: Proportion of differentially spliced clusters containing introns that are all fully annotated.

Level 3 (Diagnosis)

Figure 3.18: Proportion of differentially spliced clusters containing introns that are all fully annotated.